Research Topic: Within artificial intelligence, reinforcement learning, a sub-field of machine learning, aims to solve sequential decision problems by trial-and-error. An agent interacts with an unknown environment and learns through its own experience to maximize a reward function. Unlike supervised or unsupervised learning, the database is built online according to the decisions made by the agent. Moreover, the reward function is often much less informative than in the case of supervised learning. For instance, unlike a supervised loss, it does not have to be differentiable.
Reinforcement learning is particularity interesting to solve numerous real-world optimization problems where several linked decisions have to be made (e.g. robotics, data center control, traffic light control, etc). On the other hand, reinforcement learning can also be seen as a simulation of animal (and human) learning, and therefore could be a possible path towards artificial general intelligence.
Reinforcement Learning, Neural Networks, Transfer Learning, Developmental Learning, Machine Learning, Actor-Critic
||| Umer Siddique, Paul Weng, and Matthieu Zimmer. Learning fair policies in multi-objective deep reinforcement learning with average and discounted rewards. In International Conference on Machine Learning, 2020.
[ bib | pdf | source code ]
||| Matthieu Zimmer and Paul Weng. Exploiting the sign of the advantage function to learn deterministic policies in continuous domains. In International Joint Conferences on Artificial Intelligence, August 2019.
[ bib | poster | hal pdf | source code | slides ]
||| Matthieu Zimmer, Yann Boniface, and Alain Dutech. Developmental reinforcement learning through sensorimotor space enlargement. In The 8th Joint IEEE International Conference on Development and Learning and on Epigenetic Robotics, September 2018.
[ bib | video | hal pdf | source code | slides ]
||| Matthieu Zimmer and Stephane Doncieux. Bootstrapping q-learning for robotics from neuro-evolution results. IEEE Transactions on Cognitive and Developmental Systems, 2017.
[ bib | video | hal pdf | pdf ]